Pneumonia is the most common cause of sepsis and the most common cause of acute lung injury (ALI). The overall objective of this project is to investigate the contribution of human genetic abnormalities and bacterial virulence genes to the development of clinical lung injury and pneumonia. We have chosen a specific human genetic abnormality and a group of bacterial virulence genes (type III secretion system) to investigate because of their potential importance in lung infections, particularly lung infections in critically ill patients, and in patients at risk for ALI. Mannose binding lectin (MBL) deficiency is one of the most common immunodeficiencies and has been associated with severe lung disease due to mucosally-acquired pathogens. P.aeruginosa is the most frequent Gram negative pathogen associated with ventilator-associated pneumonia (VAP). The products of Type III secretion system genes in P.aeruginosa cause acute lung injury in experimental animals and P.aeruginosa strains isolated from critically ill patients produce type III secretion proteins. We will establish a prospective cohort investigation to test the hypothesis that genetic abnormalities in MBL are associated with an increased risk of severe ALI (Aim 1). We will also establish a prospective cohort investigation of the molecular characteristics of P.aeruginosa strains and the genetic expression of virulence genes in ventilated patients who are colonized with P.aeruginosa (Aim 2). Patients with P. aeruginosa colonization (established by surveillance cultures of tracheal aspirates) will undergo periodic BAL to determine whether colonization can be distinguished from VAP by a biochemical marker of lung epithelial injury. BAL will also be done on patients with VAP and/or ALI due to P.aeruginosa; the results of these BAL fluids will be compared to the BAL obtained from colonized patients. This aim will test the hypothesis that transformation from colonization to lung infection is associated with the expression of the Type III secretion virulence genes. We will then compare the P.aeruginosa strains obtained from patients with VAP and/or ALI to the strains obtained from patients who are only colonized with P.aeruginosa and quantify the lung injury caused by the airspace instillation of the strains into mice (Aim 3). This final aim will allow us to utilize animals to answer a clinical question; whether the colonizing strains and the strains found in patients with VAP differ in terms of their biologic effects. BAL fluids from patients and from the mice will be utilized in other projects (l&3) to evaluate the procoagulant pathway and the antifibrinolytic pathway in the distal airspaces. The fluids will also be evaluated in the proteomics core to determine whether bacterial virulence proteins can be measured in BAL fluids from patients with VAP and/or ALI and whether markers of epithelial injury can also be found in the BAL fluids of these patients.